


An Introductory Analysis of the Phonology of The Trollish Language

by 1nsomnizac



Category: Homestuck
Genre: Gen, Languages and Linguistics, Metafiction, Other, Trolls (Homestuck)
Language: English
Status: Completed
Published: 2019-04-14
Updated: 2019-11-21
Packaged: 2020-01-13 02:40:36
Rating: General Audiences
Warnings: No Archive Warnings Apply
Chapters: 2
Words: 3,839
Publisher: archiveofourown.org
Story URL: https://archiveofourown.org/works/18459776
Author URL: https://archiveofourown.org/users/1nsomnizac/pseuds/1nsomnizac
Summary: What it says on the tin. I attempt to discover what I can about the troll language by analysing the names of the trolls.





	1. Chapter 1

**Author's Note:**

> I'm probably going to clean this up later but I need to post it now. my brain is screaming.

An Introductory Analysis of the Phonology of The Trollish Language

by Zaira (1nsomnizac)

0\. Introduction

The troll characters in the webcomic Homestuck canonically speak a common language. Unfortunately for language aficionados like myself, though admittedly fortunately from a textual perspective, this language is not shown to the audience. However, the names of the troll characters seem to have a certain kinship, as though they are all words from a common language. So I thought that this April 13th, it being Homestuck's tenth anniversary, I would publish something that has been bouncing around in my head for a few years. In this paper I will attempt to reconstruct as much as possible the phonology or sound system of the hypothetical Trollish language. This is not intended to be a scholarly analysis, I do not actually know how one would go about trying to reconstruct the phonology of a language. But this is rigorous enough for a fan work for my personal satisfaction. I hope that you find it enjoyable. Happy 4/13!

1\. Starting Assumptions

  1. I assume that Trollish is an oral language. In an oral language, a speaker makes sounds which we call _phones_ with their vocal track (throat, mouth, and nose), which listeners can recognize. Generally, several similar phones are interchangeable in a given oral language. We call a group of phones which speakers interpret as “the same sound” a _phoneme_. In oral languages, speakers pronounce phonemes in bursts which we call _syllables_. Oral languages have rules that limit which sounds may go together in a syllable, and how they may go together. The nucleus (the long and loud middle part) of every syllable is a vowel. The _onset_ (beginning of a syllable) and _coda_ (end of a syllable) may be absent, or they may contain one or more consonant sounds. All human languages have both vowel phonemes and consonant phonemes, and all human languages have more consonant phonemes than vowel phonemes.
  2. I assume that the names of the trolls in Homestuck follow the rules of Trollish phonology.
  3. I assume that each letter used to write Trollish words represents an individual phoneme, until evidence shows otherwise. In English, we often use multiple letters to represent one sound, but this is mainly a product of using a writing system that was not designed for our language. 
  4. I assume that a syllable begins where a word begins, and that a syllable ends where a word ends.
  5. I assume that any vowel or uninterrupted sequence of vowels might be the nucleus of a syllable, that any consonant or series of consonants that comes before a vowel might be the onset of a syllable, and  that any consonant or series of consonants that comes after a vowel might be the coda of a syllable.
  6. I assume that the first vowel or series of vowels is the nucleus of the first syllable of the word, and that any consonants that come before it are the onset of the first syllable. Likewise, I assume that the last vowel or series of vowels is the nucleus of the last syllable of the word, and that any consonants that come after it are the coda of the first syllable.
  7. I assume that if a nucleus can take an onset in a syllable in one word, and take a coda in a syllable in another word, that the nucleus can take both the onset and the coda in the same syllable.
  8. I assume that the words of the sample reflect the normal shapes of Trollish words.
  9. I am going to assume that /a/, /e/, /i/, /o/, /u/, and /y/ are vowels and that everything else is a consonant, because otherwise we won’t get anywhere.



2\. The Sample

Below are all the Troll names given in the text of Homestuck. I am aware there are others in Hiveswap but I'm just doing these.

/ampora/, /aradia/, /aranea/, /captor/, /cronus/, /damara/, /equius/, /eridan/, /feferi/, /gamzee/, /horuss/, /kanaya/, /kankri/, /karkat/, /kurloz/, /latula/, /leijon/, /makara/, /maryam/, /meenah/, /megido/, /meulin/, /mituna/, /nepeta/, /nitram/, /peixes/, /porrim/, /pyrope/, /rufioh/, /serket/, /sollux/, /tavros/, /terezi/, /vantas/, /vriska/, /zahhak/,

3\. Initial Data

/ampora/

  * phonemes: /a/, /m/, /o/, /p/, /r/
  * onset + nucleus: /a/
  * nucleus + coda: /a/



 /aradia/

  * phonemes: /a/, /d/, /ia/, /r/
  * onset + nucleus: /a/
  * nucleus + coda: /ia/



/aranea/

  * phonemes: /a/, /ea/, /n/, /r/
  * onset + nucleus: /a/
  * nucleus + coda: /ea/



/captor/

  * phonemes: /a/, /c/, /o/, /p/, /r/, /t/
  * onset + nucleus: /ca/
  * nucleus + coda: /or/



/cronus/

  * phonemes: /c/, /n/, /o/, /r/, /s/, /u/
  * onset + nucleus: /cro/
  * nucleus + coda: /us/



/damara/

  * phonemes: /a/, /d/, /m/, /r/
  * onset + nucleus: /da/
  * nucleus + coda: /a/



/equius/

  * phonemes: /e/, /q/, /s/, /uiu/
  * onset + nucleus: /e/
  * nucleus + coda: /uius/



/eridan/

  * phonemes: /a/, /d/, /e/, /i/, /n/, /r/
  * onset + nucleus: /e/
  * nucleus + coda: /an/ 



/feferi/

  * phonemes: /e/, /f/, /i/, /r/
  * onset + nucleus: /fe/
  * nucleus + coda: /i/



/gamzee/

  * phonemes: /a/, /ee/, /g/, /m/, /z/
  * onset + nucleus: /ga/
  * nucleus + coda: /ee/



/horuss/

  * phonemes: /h/, /o/, /r/, /s/, /u/
  * onset + nucleus: /ho/
  * nucleus + coda: /uss/



/kanaya/

  * phonemes: /a/, /aya/, /k/, /n/
  * onset + nucleus: /ka/
  * nucleus + coda: /aya/



/kankri/

  * phonemes: /a/, /i/, /k/, /n/, /r/
  * onset + nucleus: /ka/
  * nucleus + coda: /i/



/karkat/

  * phonemes: /a/, /k/, /r/, /t/
  * onset + nucleus: /ka/
  * nucleus + coda: /at/



/kurloz/

  * phonemes: /k/, /l/, /o/, /r/, /z/
  * onset + nucleus: /ku/
  * nucleus + coda: /oz/



/latula/

  * phonemes: /a/, /l/, /t/, /u/
  * onset + nucleus: /la/
  * nucleus + coda: /a/



/leijon/

  * phonemes: /ei/, /j/, /l/, /n/, /o/
  * onset + nucleus: /lei/
  * nucleus + coda: /on/



/makara/

  * phonemes: /a/, /k/, /m/, /r/
  * onset + nucleus: /ma/
  * nucleus + coda: /a/



/maryam/

  * phonemes: /a/, /m/, /r/, /ya/
  * onset + nucleus: /ma/
  * nucleus + coda: /yam/



/meenah/

  * phonemes: /a/, /ee/, /h/, /m/, /n/
  * onset + nucleus: /mee/
  * nucleus + coda: /ah/



/megido/

  * phonemes: /d/, /e/, /g/, /i/, /m/, /o/
  * onset + nucleus: /me/
  * nucleus + coda: /o/



/meulin/

  * phonemes: /eu/, /i/, /l/, /m/, /n/
  * onset + nucleus: /meu/
  * nucleus + coda: /in/



/mituna/

  * phonemes: /a/, /i/, /m/, /n/, /t/, /u/
  * onset + nucleus: /mi/
  * nucleus + coda: /a/



/nepeta/

  * phonemes: /a/, /e/, /n/, /p/, /t/
  * onset + nucleus: /ne/
  * nucleus + coda: /a/



/nitram/

  * phonemes: /a/, /i/, /m/, /n/, /r/, /t/
  * onset + nucleus: /ni/
  * nucleus + coda: /am/



/peixes/

  * phonemes: /e/, /ei/, /p/, /s/, /x/
  * onset + nucleus: /pei/
  * nucleus + coda: /es/



/porrim/

  * phonemes: /m/, /o/, /p/, /r/
  * onset + nucleus: /po/
  * nucleus + coda: /im/



/pyrope/

  * phonemes: /e/, /o/, /p/, /r/, /y/
  * onset + nucleus: /py/
  * nucleus + coda: /e/



/rufioh/

  * phonemes: /f/, /h/, /io/, /r/, /u/
  * onset + nucleus: /ru/
  * nucleus + coda: /ioh/



/serket/

  * phonemes: /e/, /k/, /r/, /s/, /t/
  * onset + nucleus: /se/
  * nucleus + coda: /et/



/sollux/

  * phonemes: /l/, /o/, /s/, /u/, /x/
  * onset + nucleus: /so/
  * nucleus + coda: /ux/



/tavros/

  * phonemes: /a/, /o/, /r/, /s/, /t/, /v/
  * onset + nucleus: /ta/
  * nucleus + coda: /os/



/terezi/

  * phonemes: /e/, /i/, /r/, /t/, /z/
  * onset + nucleus: /te/
  * nucleus + coda: /i/



/vantas/

  * phonemes: /a/, /n/, /s/, /t/, /v/
  * onset + nucleus: /va/
  * nucleus + coda: /as/



/vriska/

  * phonemes: /a/, /i/, /k/, /r/, /s/, /v/
  * onset + nucleus: /vri/
  * nucleus + coda: /a/



/zahhak/

  * phonemes: /a/, /h/, /k/, /z/
  * onset + nucleus: /za/
  * nucleus + coda: /ak/



3\. Initial Analysis

The sample attests the following onset-nucleus sequences.

  * Nucleus /a/: /a/, /ca/, /da/, /ga/, /ka/, /la/, /ma/, /ta/, /va/, /za/
  * Nucleus /e/: /e/, /fe/, /me/, /ne/, /se/, /te/
  * Nucleus /ee/: /mee/
  * Nucleus /ei/: /lei/, /pei/
  * Nucleus /eu/: /meu/
  * Nucleus /i/: /mi/, /ni/, /vri/
  * Nucleus /o/: /cro/, /ho/, /po/, /so/
  * Nucleus /u/: /ku/, /ru/
  * Nucleus /y/: /py/



The sample attests the following nucleus-coda sequences.

  * Nucleus /a/: /a/, /ah/, /ak/, /am/, /an/, /as/, /at/
  * Nucleus /aya/: /aya/
  * Nucleus /e/: /e/, /es/, /et/
  * Nucleus /ea/: /ea/
  * Nucleus /ee/: /ee/
  * Nucleus /i/: /i/, /im/, /in/
  * Nucleus /ia/: /ia/
  * Nucleus /io/: /ioh/
  * Nucleus /o/: /o/, /on/, /or/, /os/, /oz/
  * Nucleus /u/: /us/, /uss/, /ux/
  * Nucleus /uiu/: /uius/
  * Nucleus /ya/: /yam/



From these attested sequences, I deduce that the following onset-nucleus-coda sequences are valid Trollish syllables.

Nucleus /a/: 

  * /a/, /ah/, /ak/, /am/, /an/, /as/, /at/ 
  * /ca/, /cah/, /cak/, /cam/, /can/, /cas/, /cat/ 
  * /da/, /dah/, /dak/, /dam/, /dan/, /das/, /dat/ 
  * /ga/, /gah/, /gak/, /gam/, /gan/, /gas/, /gat/ 
  * /ka/, /kah/, /kak/, /kam/, /kan/, /kas/, /kat/ 
  * /la/, /lah/, /lak/, /lam/, /lan/, /las/, /lat/ 
  * /ma/, /mah/, /mak/, /mam/, /man/, /mas/, /mat/ 
  * /ta/, /tah/, /tak/, /tam/, /tan/, /tas/, /tat/ 
  * /va/, /vah/, /vak/, /vam/, /van/, /vas/, /vat/ 
  * /za/, /zah/, /zak/, /zam/, /zan/, /zas/, /zat/



Nucleus /e/: 

  * /e/, /es/, /et/ 
  * /fe/, /fes/, /fet/ 
  * /me/, /mes/, /met/ 
  * /ne/, /nes/, /net/ 
  * /se/, /ses/, /set/ 
  * /te/, /tes/, /tet/



Nucleus /ee/:

  * /mee/



Nucleus /i/: 

  * /mi/, /mim/, /min/ 
  * /ni/, /nim/, /nin/ 
  * /vri/, /vrim/, /vrin/



Nucleus /o/: 

  * /cro/, /cron/, /cror/, /cros/, /croz/
  * /ho/, /hon/, /hor/, /hos/, /hoz/
  * /po/, /pon/, /por/, /pos/, /poz/
  * /so/, /son/, /sor/, /sos/, /soz/



Nucleus /u/: 

  * /kus/, /kuss/, /kux/ 
  * /rus/, /russ/, /rux/



Already, we can analyse some of the words in the sample in syllables.

  * /ampora/ = /am.por.a/
  * /horuss/ = /ho.russ/
  * /vantas/ = /van.tas/



We can see the following syllable patterns on display.

  * Vowel
  * Consonant + Vowel
  * Consonant + Vowel + Vowel
  * Consonant + Vowel + Consonant
  * Consonant + Vowel + Consonant + Consonant
  * Consonant + Consonant + Vowel



All human languages allow V and CV syllables. Linguists represent this structure as (C)V - optional consonant plus vowel. Based on these syllable patterns, we might conclude that the Trollish syllable pattern is (C)(C)V(V)(C)(C). However, there are a few important caveats we should keep in mind.

First, the only onsets that contained two consonants were /cr/ and /vr/. Both could be more conservatively described as (C)(r). We do not know what consonant sound /c/ represents, so we cannot be any more conservative in our analysis (we want to have as solid a groundwork as possible now, so that future discoveries will only upset the later stages of our analysis).

The letter /r/ represents one of a number of different sounds. Linguists have different symbols for the common Spanish R, the relatively rare American R, the throaty French R, and the Japanese tapped R. All of these sounds are what linguists call  _ approximants _ , sounds which we make by only gently restricting the flow of air from the lungs through our mouth. Many languages, such as Japanese, avoid clusters of consonants like SK or FR, but gladly accept words with a consonant + approximant pattern, such as  _ Kyoto _ ,  _ Ryukyu _ , or  _ nyaa _ . It is quite possible that Trollish would allow a syllable like /vri/ but not a syllable like /spa/.

Next, we are not sure which sequences of vowel letters have one syllable, and which have many. Since we do not want to speculate just yet, let’s just keep our analysis as (C)(r)V for now.

Third, the only coda with two consonants is /ss/. In some languages, a double consonant represents a “geminated” or doubled consonant, a consonant held for twice the length of a normal one. If you speak Finnish or Estonian you are already familiar with this concept; the difference between the Finnish  _ taka  _ and  _ takka  _ is the difference between  _ back  _ and  _ fireplace _ . English uses gemination in a few places; we say misspell with a geminated S (mis-spell rather than mis-pell), roommate rather than roomate. Geminated consonants usually occur between vowels, but some languages, like Punjabi, allow word final geminates. It is quite possible that this /ss/ represents a geminate /s/ sound. 

However, it is also possible that the second /s/ instead signals a feature of the syllable, like double consonants often do in Germanic and Romance languages. In these languages, a double letter indicates that the preceding vowel has some unusual feature. Usually it means that the vowel is short, in other words, pronounced for a shorter period of time, and often slightly differently. In English, it only represents a change in vowel quality, since length is not a feature of English words. It might also signal an unusual placement of a stressed syllable.

For now, I can only guess what a double /s/ means. So for now, I am going to ignore it and focus on firmer territory. 

4\. Results of Initial Analysis

I propose the syllable structure (C)(r)V(C), with the caveat that /r/ may only follow stops, affricates, and fricatives.

I explained what an approximant was above. Two approximants in a row is unusual in onsets, though technically possible. A  _ fricative  _ is a sound that we make by restricting the flow of air from our mouths to a narrow stream. The English sounds F, V, S, Z, SH, soft J, and hard and soft TH are fricatives. A stop, as the name implies, is a sound that you make by temporarily stopping the flow of air from your mouth. The English P, B, T, D, K, and hard G are all stops. An affricate is a sound that starts as a stop but ends as a fricative, like the English CH and hard J sounds (CH starts as T and becomes SH, hard J starts out D and becomes soft J).

Since we know that /vr/ is a valid onset, it would make sense to assume that /fr/ is also valid. /sr/ and /zr/ may also be possible, even though these are unusual to an English speaker. We do not know whether /c/ is a fricative or not. If it is, then we could narrow our possible onsets down to fricative + /r/. It is highly likely that /c/ is a stop, or even . Therefore, we leave the identity of /c/ unknown and move on. 

I posit that the following are valid onsets.

  * Nasal: /m/, /n/
  * Stop: /d/, /g/, /k/, /p/, /t/
  * Fricative: /f/, /s/, /v/, /z/
  * Approximant: /h/, /l/, /r/
  * Unknown consonant type: /c/
  * Stop/Fricative/Unknown + /r/: /cr/, /dr/, /fr/, /gr/, /kr/, /pr/, /sr/, /tr/, /vr/, /zr/



I posit that the following are all valid nucleuses.

  * Vowels: /a/, /e/, /i/, /o/, /u/, /y/



I posit that the following are valid codas:

  * Nasal: /m/, /n/
  * Stop: /d/, /g/, /k/, /p/, /t/
  * Fricative: /f/, /s/, /v/, /z/
  * Approximant: /h/, /l/, /r/
  * Unknown consonant type: /x/



Given these structures, I can analyse the syllables of most of the sample as follows. The highlighted syllables are the ones which I confirmed earlier in my analysis.

  * /ampora/ = / **am**. **por**. **a** /
  * /aradia/ = / **a**.ra.di. **a** /
  * /aranea/ = / **a**.ra.ne. **a** / 
  * /captor/ = /cap.tor/
  * /cronus/ = / **cro**.nus/
  * /damara/ = / **da**. **ma**.ra/
  * /eridan/ = / **e**.ri. **dan** /
  * /feferi/ = / **fe**. **fe**.ri/
  * /kankri/ = / **kan**.kri/
  * /karkat/ = /kar. **kat** /
  * /kurloz/ = /kur.loz/
  * /latula/ = / **la**.tu. **la** /
  * /makara/ = / **ma**. **ka**.ra/
  * /megido/ = / **me**.gi.do/
  * /mituna/ = / **mi**.tu. **na** /
  * /nepeta/ = / **ne**.pe. **ta** /
  * /nitram/ = / **ni**.tram/ or /nit.ram/
  * /porrim/ = / **por**.rim/
  * /pyrope/ = /py.ro.pe/
  * /rufioh/ = /ru.fi.oh/
  * /serket/ = /ser.ket/
  * /sollux/ = /sol.lux/
  * /tavros/ = / **ta**.vros/
  * /terezi/ = / **te**.re.zi/
  * /vantas/ = / **van**. **tas** /
  * /vriska/ = /vris. **ka** /
  * /zahhak/ = / **zah**.hak/



Some of these readings do not fully satisfy me. They do not really "feel" like the way that my internal sense of what the language should sound like based on these words. In addition, many words of the sample just do not really work with this structure, since they include sequences of letters which my analytical technique is not well adapted to handle.

However, I have come to the end of the section of this project where I am able to analyse the language based solely on firm data. Any further analysis will require leaps of intuition, which I do not wish to publish just yet. Further research into the Trollish language is required.


	2. Addendum: Canonical Troll Physiology And Linguistics

**Summary for the Chapter:**

> Some thoughts on the way that troll bodies may affect troll languages

Addendum, November 20, 2019

0\. Preface

Communication between people requires "signs". Signs are forms, perceivable pieces of a person's sensory input, paired with a meaning. In oral languages like spoken English, these signs are sounds that a human ear can hear. In written English, these are sequences of letters that the human eye can see. In American Sign Language, these are hand signs.

To put it another way, I need to be able to make signs that other people can perceive, and I need to be able to perceive the signs that other people make. In the valley of the blind, sign language is not likely to be popular, but in the valley of the deaf, oral language is king. Yet in both of these places, the general shape of human bodies, deaf or blind, shape the languages. There are no human sign languages that require the use of four arms, or antennae, or cheek fins, or "bone bulges," and there are no human oral languages that feature the clicking of external mandibles, or high-frequency vibrations like the ones which some cetaceans use to navigate the sea.

Even with these restrictions, however, we have to remember that humans do vary quite a bit in their ability to make signs, and we are perfectly able to create social groups whose members are not equally capable of communicating in the generally accepted way. For example, the society of the United States, my home country, is built on the assumption that everyone is able to hear, interpret, and speak English, and despite the fact that deafness, blindness, and muteness predate the genesis of our species, we have yet to create a form of language which every human can learn. We instead have an oral language, and a list of terms for people who cannot keep up. Let us call the ability to produce a sign of a language a "sign capacity".

How do trolls and humans differ? What range of sign capacities do Alternians consider "normal," and which abnormal or deficient?

1\. Nasality

Do trolls have noses? They obviously must; after all, Terezi Pyrope uses her sense of smell to navigate the world. But do they have the same kind of noses as humans?

The human head is full of an empty-ish region called the sinuses or the sinus cavities. These sinus cavities have two exits, so to speak, the double door that is the nostrils, and the back door to your windpipe through the back alley that is your oral cavity. When air passes from your lungs and out your nose, it makes a nasal sound. When you raise your velum, the soft part of the back of the roof of your mouth, you shut the back door, and your sound is no longer nasal. This is also why your voice sounds different when you have a cold or a sinus infection. The mucus that your body produces blocks air from moving from the back door to the front.

In English, three pairs of consonant phonemes are distinguished by nasality. Listen to the difference between the following pairs: dumb and dub, done and dud, dung and dug. The last consonant in each pair is distinct only because the first is nasal and the second is not.

English does not distinguish between plain vowels and nasal vowels, but plenty of other languages do. In French, the vowel in the word _vein_ (pronounced /vɛ̃/) has a nasal vowel, while the word _veine_ (pronounced /vɛn/) does not. In many languages around the world, including several prominent North American, West Africa, and Southeast Asia lean heavily on this feature.

In order to use nasality as a feature of oral language, the nasal passage and the oral cavity must be connected, and an individual must be able to control whether or not air goes through the nasal passage.

If we determine that this is the case, this means that nasal sounds are an available feature. If we assume that the /n/ and /m/ in the names of trolls which I analysed in the previous chapter refer to the same sounds they do in English, then we already know this.

However, there is a high chance that trolls have very non-human internal plumbing. The fact that some trolls can breathe underwater and apparently have gills suggests that trolls do not breathe the same way that humans do. It is highly likely that at least some trolls, particularly seadwellers, are unable to use nasal sounds, or perhaps have nasal sounds that are noticeably different from those of other trolls, or from humans.

2\. Labial and Dental Sounds

Trolls have big teeth. At least, many trolls seem to. Kanaya appears to have fangs which rest on her lips, and Sollux appears to have a noticeable speech impediment that is caused by or exacerbated by his teeth. This likely affects labial and dental sounds, at least among trolls which have overhanging teeth.

As a side note, The word labial comes from the Latin word _labium_ , meaning "lip." the word 'dental' comes from the Latin word _dens_ / _dentis_ , meaning "tooth". Thus, a labiodental sound is a sound which people produce by resting their teeth against their lip, rather than the sort of genital horror out of the movie _Teeth_ which an English speaker is likely to conjure out of the alternate meaning of the Latin word labium. I once shocked someone by referring to the Japanese F sound as a "bilabial fricative," so I thought that it would be best to clarify.

Many languages have bilabial sounds (made with both lips), labiodental sounds (teeth against lip), linguodental sounds (tongue against teeth). For example, the English words rip, rib, and rim end in bilabial sounds, the words safe and save end with labiodental sounds, and the words breath and breathe end in linguodental sounds. If your top front teeth are a row of razor sharp shark teeth like Feferi Peixes, you could really fuck up your lips and tongue whenever you say a word like "thief," and if your huge front teeth hang over your lower lip all the time, like... about half of the trolls if their sprites are to be believed, the B, P, and M sounds might not be easy sounds for you to make.

What complicates this is the fact that trolls seem to have a wide range of types of (fucked-up knife teeth) nonhuman dentition. This suggests to me that either labial and dental consonants are avoided in Trollish, or that these phonemes of Trollish may be realized in a very wide variety of ways.

This might explain two common speech excentricities. Sollux appears to have a lisp in the text, which disappears when his fangs are knocked out. But the S and Z sounds are not made with tongue-to-tooth contact. It may be that the lisp is a "translation" of an inability to make a TH, F, and/or V sound in a way that other trolls read as acceptable. Eridan's "wwobbly" speech pattern, with emphasized Ws and Vs, both of which involve lip action, might be a result of an unusually _normal_ (to a human) ability to enunciate labial sounds.

With all this difficulty with front consonants, it would not surprise me if Trollish made heavy use of "guttural" sounds (sounds made farther back in the mouth), which are less often affected by a person's teeth.

Lip rounding is a common feature in languages which might also be hindered. The vowels in the English words new, no, now, and nook all involve rounded lips. The consonants at the beginning of the words _whip_ , _chip_ , and _ship_ are also made with rounded lips. This feature is likely absent or non-contrastive, that is, there is no contrast in meaning associated with a contrast in sound.

3\. Additional Notes

This is all that I can think of for now. I may add more at a later date.


End file.
